sven

Hi people, I have a ZFS cache pool in addition to my regular HDD array, running on 6.12.6. That ZFS pool consists of two identical Samsung pm883 480gb ssd drives in a mirror-0 configuration. The two drives sit in hotswappable bays (with 4 empty slots remaining). The drives are 6 months old. One drive is slowly but consistently giving "CRC error count" errors since the beginning. Sometimes it goes up by 1, sometimes by 2 or 3, and in between there are weeks or months of 0 errors. Currently sitting at 23 after about six months. The other drive has never given an error. From what I have read, this CRC error usually goes up a lot faster in a short time if it is due to a cabling issue. I want to see if my issue is related to the cabling or port/controller. Can I just stop the array, shut the server down, and swap the two SSD's between the two bays they are currently in? And afterwards do I need to change something with the ZFS configuration or can I just start the array?

I am a new user of the File Integrity Plugin. I have used the BLAKE3 algorithm and so far the performance seems great. One (small?) problem: at the end of running the Export step for one of the disks, I get a PHP error that a memory limit has been reached: I have gathered that this has to do with the memory limit of php which is set at 128MB. When I go to the export directory, I see that a 168MB file has been created called disk1.export.blake3.hash. When I open the file it contains the hashes for all files from disk 1. Since the export file seems to be fine, is this just an error in the progress display? There were 862892 files in the export. 6TB drive, many large (500MB+) files, but also a lot of smaller files.